Nucleoside Triphosphate Pools in Synchronous Cultures of Escherichia colil

Transcription

1 JOURNAL OF BACTERIOLOGY. Oct. 1971, p Copyright 1971 American Society for Microbiology Vol. 18, No. I Printed in U.S.A. Nucleoside Triphosphate Pools in Synchronous Cultures of Escherichia colil LYDIA HUZYK AND D. JOSEPH CLARK Department of Microbiology, The University of British Columbia, Vancouver 8, British Columbia Received for publication 8 June 1971 Endogenous nucleoside triphosphate pools in synchronized cultures of Escherichia coli B/r/1 oscillate as a function of age. Purine nucleoside triphosphates show a gradual 5% increase from zero age to the time of subsequent division, immediately prior to division. In contrast, pyrimidine nucleoside triphosphates undergo a dramatic change of about 5% in the first half of the generation at a time coincident with the termination of a round of deoxyribonucleic acid replication. A 5 to 7% increase starts at the initiation of the next round of deoxyribonucleic acid replication and continues until cell division, in parallel with the purine nucleotides. The fluctuation of pyrimidines between zero age and the middle of the division cycle suggests a functional relationship for pyrimidine metabolism and the regulation of cell division. The biochemical mechanism which couples deoxyribonucleic acid (DNA) replication and cell division is not understood, although some activities of enzymes involved in nucleic acid metabolism during the cell cycle of eucaryotic cells have been examined. Johnson and Schmidt (11) found periodic changes in thymidylate kinase in synchronous cultures of Chlorella pyrenoidoso. The enzyme reached maximal activity prior to the maximal rate of DNA synthesis. Turner et al. () found a strict correlation between ribonucleotide reductase levels and that fraction of L cells which were replicating their DNA. After DNA replication, there was a continuous decay of enzymatic activity, which decreased shortly after cell division in Chinese hamster fibroblast cells (). Cleaver described the increase of thymidine enzyme activities at the onset of DNA replication and then a decrease in their activity as "DNA synthesis ceases" (4). He suggested that the close relationship of these enzymes with DNA synthesis may reflect control of initiation of DNA replication. In bacteria, there is no information about the activity of enzymes involved in nucleic acid metabolism. Lark (1) examined the fluctuation of deoxyribonucleosides as a function of the cell cycle of Salmonella typhimurium. He noted an increase in the pool of thymidine coincident with the age at which there was increased synthesis of DNA. He did not measure the purine deoxyribonucleosides. Since cell division is dependent upon DNA replication (3, 8, 13), and the end of a round of replication is required for cell division in normal cells (3, 9), it is possible that a signal for division could be activated as a result of the completion of the round of DNA replication. That signal could be directly or indirectly related to the precursors of DNA synthesis. This report describes fluctuations in the endogenous nucleoside triphosphate pools as a function of the growth cycle of Escherichia coli B/r. MATERIALS AND METHODS Bacterial strains and culture methods. E. coli B/r/l (ATCC 147), a prototroph, was used for synchronous cultures since it forms single cells. The cultures were maintained on nutrient agar slants and in liquid cultures of.5 M tris(hydroxymethyl)aminomethane (Tris)-glucose-mineral salts medium. A growth medium that was low in P4 was used to provide efficient incorporation of Spi into nucleic acids. The low P4 media had the following composition:.5 M Tris, x 1-5 M KHPO4, 1-4 M CaCI, x 1- M MgCI, 3 x 1-6 M FeCI.7H, 5. x 1- M NaCI, and 3 x 1- M (NH4)SO4 x.% carbon source. Cultures were incubated at 37 C in a shaking water bath. Synchronization of the cells was accomplished by using the technique of.helmstetter and Cummings (7). Cell growth was measured with a modified Coulter counter (3) Ḋetermination of acid-soluble nucleoside triphospbate 1 This report was taken from a dissertation submitted by L. H. to the University of British Columbia Gallant (1). Chromatographic separation was carried pools. Samples were prepared by the method of Irr and in partial fulfillment of the requirement for the M.S. out on polyethyleneimine (PEI)-impregnated cellulose degree (197). thin layers prepared by the method of Randerath and 74

2 VOL. 18, 1971 NUCLEOTIDE POOLS IN E. COLI SYNCHRONOUS CELLS Randerath (18, 19). Preliminary washings of the thin layers and chromatography were carried out by the methods of Irr and Gallant (1). Slight modifications were made in the solvents. In the first direction, the first solvent was I N acetate-l M LiCl (1:1, v/v) for 1 cm; the second solvent was 1 N acetate-1.5 M LiCl (1: 1, v/v) for the remainder of 5.5 hr. The solvent of the second direction was 3 M NH4 acetate plus 4.3% borate (ph 7) for 4.5 hr. To ensure distinct separation of deoxyribonucleotides from their corresponding ribonucleotides, wicks made of Whatman no. 3 paper were attached to the thin layer for each direction. Autoradiography and removal of nucleoside triphosphates were done by the methods of Irr and Gallant (1). The film was exposed to the radioactive chromatogram for 3 days. Counts per minute were corrected for isotope decay, counting efficiency, volume spotted, conversion factor for microcuries to disintegrations per minute, specific activity, and rate of mass increase to give nanomoles of 3Pi per milliliter per nucleotide. Chemicals. 3Pi (orthophosphate-free) was obtained from Tracerlab; 3H-thymidine was from Schwartz Bio Research, Inc.; unlabeled marker nucleotides were from Calbiochem; PEI was from Chemirad Corp.; and Cellulose MN 3 HR was from Canlab. RESULTS Growth curves in low phosphate concentrations. The concentration of deoxynucleotide and ribonucleotide precursors in bacteria is extremely low 1. I I. I. 8_ - 6D 75 (14). Since the yield of cells obtained for synchronous cultures is quite low ( x 16 to 1 x 17 cells per ml) when the membrane technique for synchronization is used, a high specific activity of SPi is required to obtain sufficient labeling of the precursors. Since 3Pi emits beta rays which are deleterious for biological systems, a specific activity of 3Pi must be obtained high enough to measure the pools yet low enough in total activity so as not to disturb the physiology of the cells. The optimum concentration of P4 and the amount of 8pi to be added to the cultures were determined by using a Coulter counter to monitor exponential growth of cell numbers and the size distribution of the population. A shift in the size distribution of a population was reflected by a change in the peak position of the size distribution curve. Such a change is considered an index of unbalanced growth (19). As seen in Fig. 1, a PO concentration of I x 1-4 M allowed an exponential increase in cell number to a density of 7.5 x 18 cells per ml; a shift in the size distribution, however, occurred at 5 x 18 cells per ml. Similar analyses were applied to cultures growing at x 1-5 and 5 x l-6 M PO4. Since x 1- M PO showed no change in balanced growth until a density of x 18 cells per ml was reached (Fig. ), this concentration 1 4_ E 1 _ c 8-/ ~~~~~ -s6 - / M inutes FIG. 1. Effect of PO4 concentrations on the growth of Escherichia coli B/ril. Cells grown for several generations in 1-4 M P4 (), S x 1-5 M PO4 (), or x 1-5 M PO4 (e) were followed with a Coulter counter for several hours. The increase in cell number is plotted as a function of time. The arrow (1) indicates the time at which a shift in the size distribution of the total population occurs.

3 76 HUZYK AND CLARK J. BACTERIOL. of PO was chosen as the lower limit for subsequent experiments. Exponential increases of cell numbers continued almost to 3 x 18 cells per ml. Figure shows a shift in the size distribution of a population limited for P,. Effect of different amouts of Pi. Growth curves based on viable cell number were constructed at different specific activities of 8pi for cells which were grown in the same concentration of PO, ( x 1-5 M PO4). A culture which had been growing for six generations and had reached a density of 17 cells per ml was diluted into culture vessels containing the following specific activities of 3pi: (i) 1 1Ci of 3Pi/Mmole of PO,; (ii) 5 gci of 3Pi/,Mmole of PO; (iii) 5,Ci of 3Pi/,Mmole of PO; and (iv) a control with no radioactivity. At regular time intervals, samples were diluted and plated for viable counts. Figure 3 shows the growth rate obtained by plotting viable cells as a function of time. The doubling time for the cultures containing 1 and 5 ACi of 3Pi/mole of PO is identical to the control. On the other hand, the culture containing 5 IACi of 3Pi/jumole of PO showed a loss in viability after 6 min of incubation. This experiment was repeated by using optical density readings at 66 nm over a 3-hr period; the results were similar to those obtained when viability was measured. Thus the cell death in the presence of a specific activity of 5 uci of 3Pi/Mmole of PO4 was due to an immediate effect and not to some delayed action of the radioactive 3Pi affecting a small clone of bacteria on a plate. A specific activity of 1 MCi of 3Pi/Mmole of PO was chosen as the optimal working level. In an effort to alleviate the deleterious effects of 3Pi, similar experiments were repeated with 33Pi. This isotope has a lower energy level (.5 FIG.. Shift in the size distribution of Escherichia coli as the culture stops growing in x 1-6 M PO4. The designations 4 and 57 indicate the time at which the size distribution is monitored for cells grown in x 1-6 M PC4 (@) from Fig. 1, and describes the size distribution of normal cells and cells starved of PC4, respectively. Ordinate refers to relative cell number; abscissa is increasing cell size. co.. C E. 7a M Inutes FIG. 3. Effect of the specific activity of 3Pi/PO4 on the growth of Escherichia coli in x 1-' M PO4 as indicated by viable cell number. Cultures grown for several generations in x 1- M PO4 received 3Pi which gave a -final specific activity of 1 MCi of 3Pi/Mmole of P4 (-); 5 MCi of 3Pi/umole of PO4 (1O); or 5 MCi of 3Pi/gmole of PO4 (e). No radioactivity was added to a control culture (). Viable cell number is plotted as a function of time. Mev, cf. 1.7 for 3Pi) and a longer half-life (5 days, cf for 3Pi; New England Nuclear Corp.). Presumably 33Pi is safer to handle and would allow longer-term experiments. Much to our surprise, 33Pi at the same specific activities used for 3Pi showed loss of viability after only a few minutes in 1 MCi of 33Pi/Mmole of PO, and the cells at higher specific activities showed no viability at all. Levels of endogenons nucleoside tnphosphate pools in exponential cels. Control levels of nucleoside triphosphate pools were established in exponentially growing cultures of E. coli B/r/1 grown in a turbidostat. A sample rate was predetermined, and the culture volume was adjusted so that dilution rate would correspond to the specific growth rate (.91/hr). A culture was grown for several generations to a density of 18 cells/ml, at which time 8pi was added to give a specific activity of 1 MCi of 3Pi/Mmole of PO, and dilution of the culture was started to maintain the steady state. The culture was di-

4 VOL. 18, 1971 NUCLEOTIDE POOLS IN E. COLI SYNCHRONOUS CELLS 77 luted with medium of the same PO, concentration and of the same specific activity of 3Pi/POQ. In this manner, an exponential culture was maintained at 11 cells per ml in the presence of 1 MUCi of 3Pi/Mmole of PO. After a 3-min equilibration period, samples were withdrawn every 5 min for hr and analyzed as described previously. A balanced steady state was evidenced by radioactive counts per minute per nucleotide which were almost identical in all samples. An example of the degree of fluctuation is illustrated in Fig. 4 for two of the eight nucleoside triphosphates, deoxythymidine triphosphate (dttp) and uridine triphosphate (UTP). The highest fluctuation for any nucleoside triphosphate was less than 1%. The actual concentrations of the endogenous nucleoside triphosphates are listed in Table 1. Nucleouide tposphate pools as a fu.tiou of the division cycle (cell age). Figure 5 shows the growth curve of a synchronous culture which was analyzed for nucleoside triphosphate concentrations. Also shown in Fig. 5 are three different TABLE 1. Nucleoside triphosphate pool levels in exponentially growing Eschenichia coli B/ril Nucleoside triphosphatea Amt of *pjb GTP.1 ATP.559 CTP.151 UTP.153 dgtp. datp.48 dctp. dttp..19 a Abbreviations: GTP, guanosine triphosphate; ATP, adenosine triphosphate; CTP, cytidine triphosphate; UTP, uridine triphosphate; d, deoxy. b Expressed as nanomoles of Jp; per nucleotide per 16 cells per milliliter. co so X 1 to) 8 X 6.84 E i_ I I I I I I I %, O M inutes FIG. 4. Level of uridine triphosphate (UTP) and deoxythymidine triphosphate (dttp) as established by measurements of 3Pi incorporation during the steadystage growth of Escherichia coli in a turbidostat. Cells were maintained at 18 cells per ml in x 1-5 M Po4 with a specific activity of 1 MCi of 3Pi/ umole of P4. The concentration of 3Pi incorporated is plotted as a function of time. Similar measurements were done for all of the nucleoside triphosphates; UTP and dttp are given as examples. A summary is presented in Table a I 1 I I I. a I Cell a-g (minutes) FIG. 5. Relationship between the increase in mass and the increase in cell number during synchronous growth of Escherichia coli. The increase in cell number during synchronous growth in x 1-5 M P4 is plotted as function of time (). The increase in cell mass is plotted as a function of time assuming: (i) an exponential increase in cell volume (solid line); (ii) a linear increase in cell volume (dashed line); or (iii) accelerated growth just prior to cell division (dotted line). curves, drawn to illustrate three possible types of mass increases during synchronous growth: linear, exponential, and a slow to rapid increase in growth rate. Since nucleoside triphosphates are expressed as concentrations, corrections are necessary on a per mass basis rather than on a per number basis. Nucleoside triphosphate concentrations were corrected assuming the Painter- Marr model for mass increase (16), in which cells exhibit the slowest rate of mass increase immediately after cell division, and those just beyond a half generation grow at the maximal

5 78 HUZYK AND CLARK J. BACTERIOL. rate. There is supporting evidence for this model from time-lapse photography studies (1) of single cells of E. coli. The synchronous culture was grown for one generation in the presence of 3pi before sampling for analysis of nucleotides. Samples were taken at 3-min intervals during the period of the second generation of growth, and were analyzed for nucleoside triphosphates as described above. The fluctuation in the nucleotide levels is indicated in Fig. 6, during a full generation of 4 min. Bearing in mind that the data have been corrected for mass increase, the purine nucleoside triphosphates increased in concentration, reached a maximum at 75 min, and dropped sharply as cell division occurred (Fig. 6A and 6C). There was an exaggerated effect on deoxyadenosine triphosphate (datp), but clearly the maximal concentration of purine nucleoside triphosphates occurred between 6 and 75 min. In all cases, there was approximately a 5% increase in the concentration near the time of cell division as compared to the levels at zero time (4 min). The pyrimidine nucleotides all exhibited a double-peak pattern during the division cycle between the time of cell division and termination of DNA replication, and again just before the following cell division (Fig. 6B and 6D). The purine deoxyribonucleoside triphosphate concentrations showed almost parallel fluctuation patterns (Fig. 6C), whereas the pyrimidine ribonucleotides, cytidine triphosphate (CTP) and UTP, oscillated in opposite directions (Fig. 6B). The information in Fig. 6 was computed according to percentage of change and is presented in Fig. 7A and B. Purine ribo- and deoxyribonucleoside triphosphates are shown in Fig. 7A. In Fig. 7B, the data for pyrimidine ribo- and deoxyribonucleoside triphosphates are presented in a similar manner. DISCUSSION In our examination of endogenous nucleoside triphosphates during the growth cycle of E. coli, we made a concerted effort to maintain balanced growth of a synchronous culture. Figure 3 indicates that a concentration of x lo-" M P4 maintains balanced growth to x 18 cells per ml. Since our cell density never exceeded 1 x 18 cells per ml, we felt that x 1-6 M PO offered sufficient margin of safety that the cells would not be adversely affected by the concentration of PO we emploved. The effects of increasing the specific activity indicated that a specific activity of 1 and 5 pci of 8Pi per,mole of PO did not adversely affect the growth of the culture as determined by viable counts and by the size distribution as a function of time (Fig. 4). On the other hand, the viability drops sharply when the cells have grown for just over one generation in the presence of 5 gci of 3pi per pmole of PO (Fig. 4). By combining these two results (Fig. 3 and Fig. 4), the lower limits of PO concentration and the upper limit of the specific activity of 3Pi per,umole of PO were established at x 1-6 M PO and 1,uCi Of 3pi per umole of PO, respectively. From the summary of the steady-state concentrations given in Table 1, the concentration of ribonucleoside triphosphates is found to be very nearly 1 times greater than their deoxyribonucleotide counterparts. The concentration of the ribonucleoside triphosphates decrease in the order adenosine triphosphate (ATP), guanosine triphosphate (GTP), and CTP = UTP. Likewise, the concentration of the deoxyribonucleoside triphosphates decreases in the order datp, dgtp, and dctp = dttp. These results are in agreement with those of others (5, 14). The concentrations of nucleoside triphosphates as a function of the cell cycle are shown in Fig. 6A and 6B. The percentage of change of the purine and pyrimidine nucleoside triphosphates is given in Fig. 7A and 7B, respectively. It is clear that both purines and pyrimidines increase just before cell division. On the other hand, dttp and UTP also increase at or about the middle of the division cycle. There is a slight increase in dctp about the same time, but it is not clear that the increase is significant. Alternatively, dctp decreases just slightly after the increase in dttp. Clearly, the decrease in dctp must occur from either decreased synthesis or increased conversion of dctp. The increase in dttp could occur due to an increased rate of synthesis or a decreased rate of conversion of dttp to the product (presumeably DNA). One of the combinations which seems attractive to us is a reduced conversion of dttp by incorporation into DNA as rounds of DNA replication are completed. This is sufficient to explain the increase in dttp but does not explain the decrease in dctp and CTP (Fig. 7B). An increase in dttp could account for a concomitant decrease in CTP and dctp, since low dttp generally is correlated with an increase in CTP and dctp (, 16, 17). It has been shown that dttp is a universal activator of ribonucleoside diphosphate reductase (). Increased activity of the reductase would decrease the level of CTP, as observed in Fig. 7B. If CTP is rapidly converted to dctp, one would anticipate an immediate increase in dctp. As CTP decreases, one would anticipate that dctp would eventually decrease as the precursor (CTP) supply is ex-

6 .4 A' ofo.4 r C I I I E 1' o CL a..35 ~ATP ~~~ 1 S C ec I * I I Z Cel I age (minutes) GTP.35.3 E.5 tl N.o E J6 4 ~~ J datp S, I I I I dgtp Cell age (minutes).4 t 4. * *l o *' : B 1 5 _ o. UTP,". O-, O.b.-o-, CTP _ )Di 4 5 6D 7 8 Cell age (minutes) I E N. 1 CL Cell age (minutes) FIG. 6. Concentration of deoxy- and ribonucleoside triphosphates during synchronous growth of Escherichia coli B/r/l. Synchronous cultures were grown at a density of approximately x 17 cells per ml in x 1-5 M P4 with a specific activity of 1 MCi of 3Pi/umole ofpo4 Samples were taken from 45 to 9 min, which corresponds to the beginning and to the end of the second cell division cycle. A, Fluctuation of purine ribonucleotides A TP () and GTP (). B, Fluctuation ofpyrimidine ribonucleotides UTP () and CTP (). C, Fluctuation of purine deoxyribonucleotides da TP () and dgtp (). D, Fluctuation ofpyrimidine deoxyribonucleotides dctp () and dttp (). Abbreviations used in Fig. 6 and 7: ATP, adenosine triphosphate; GTP, guanosine triphosphate; CTP, cytidine triphosphate; datp, deoxyadenosine triphosphate; dgtp, deoxyguanosinetriphosphate; dctp, deoxycytidine triphosphate; dttp, deoxythimidine triphosphate. 79

7 8 HUZYK AND CLARK J. BACTERIOL. ig z I.g 9 z w ct..i (L +1 t5-5 _datrp GTP dgtp -1 A I MINUTES +'I +5~ o -S4 UT 4\ dttp dctp " 'tiv -I IV -1 IB I MINUTES FIG. 7. Data from Fig. 6 converted to percentage of increase or decrease relative to the first sample taken at 45 ntin and plotted as a function of the cell age of the culture. A, Percentage of increase or decrease ofpurine ribo- and deoxyribonucleoside triphosphates; A TP (dash-and-dot line), GTP (dotted line), datp (dashed line), and dg TP (solid line). B, Percentage ofincrease or decrease of pyrimidine nucleoside triphosphates; CTP (dashed line), UTP (solid line), dctp (dash-and-dot line), and dttp (dotted line). hausted, providing dctp is continuously converted to dutp. This seems to be exactly the case as seen in Fig. 7B. As CTP decreases coincident with the increase in dttp, there is a short burst of dctp and then the dctp concentration falls to a minimum after the decrease in CTP. The decrease in CTP with a decrease in dctp following shortly thereafter, implies that the conversion of UTP to CTP is blocked. There is no known mechanism for blocking CTP synthetase which is compatible with our results, but such a block would explain the accumulation of UTP and the depletion of CTP seen in the middle of the cell cycle (Fig. 7B). Lark (1) observed an increase in thymidine coincident with an increase in DNA synthesis. Although we have observed an increase in dttp at or near the start of a round of replication, we think our results are supportive, in spite of the fact that Lark examined nucleosides and we examined nucleotides. From additional experiments, Lark concluded that there was no causal relationship between the initiation of DNA replication and the fluctuation in nucleoside pools. From our experiments, we cannot establish any causal relationship between the fluctuation of the nucleotide pools and the initiation of DNA replication. Indeed, at this growth rate (generation time 4 min), the end of the round of DNA = replication and the beginning of a new round of replication both occur at a cell age of approximately min. Therefore, it is impossible to correlate the increase in dttp with one or the other. To make such a conclusion, additional experiments should be made with generation times in which the ends of the rounds at the initiation of DNA replication are not coincident. Recent experiments by Werner (3) cast some suspicion upon the actual precursor of DNA synthesis. His experiments are interpreted to mean that deoxythymidine monophosphate (dtmp), and not dttp, is the immediate precursor of DNA replication. If so, one can legitimately question the relevancy of fluctuations in the deoxyribonucleoside triphosphates to the replication cycle of DNA. We can only state: (i) that there is a dramatic fluctuation in the nucleoside triphosphates over the cell division cycle (Fig. 7A and Fig. 7B); and (ii) whether or not dttp is the immediate precursor of DNA, the fluctuations in dttp may be indirectly, but nevertheless precisely, related to the replication cycle. If a dependence relationship exists between DNA replication and cell division, one would anticipate that cell division is dependent upon the completion of DNA replication. Clearly, there is a severe penalty for cells which divide in the absence of DNA synthesis, but it is not obvious that cells which continue to replicate DNA in the absence of cell division, i.e., those cells which form filaments, are at any severe disadvantage. It seems logical that the coordination of cell division and DNA replication would be mediated by a mechanism which has its origin and its dependence upon the state of DNA replication. Experimental data support this interpretation (3). The nucleotide precursor for DNA synthesis might provide a biochemical mechanism by which DNA replication and cell division are coupled. Since dttp is unique to DNA, it or a closely related compound might serve as the signal to coordinate cell division and DNA replication. Upon completion of DNA replication, the reduction in growing point would lead to an accumulation of dttp or a closely related compound. Increase to an active threshold level would activate those division processes required to effect or to complete cell division. This cannot be a sufficient explanation, since dttp fluctuates twice during the cell cycle. On the other hand, the two situations are distinct in that the ratio of all nucleoside triphosphates is constant near division but is vastly different at or near the end of the round. Thus, a combination of, say, high dttp and low dctp at the end of a round of DNA replication may be required as the signal for cell division. In fact, the condition of low dctp or CTP which occurs only at the end of the round might be the controlling effector. Since cyclic AMP (6, 17) and ppgpp (1) appear to interact with the CAP protein and the Psi factor for the ribosomal locus, respectively, we find it tempting to suggest that dctp or

8 VOL. 18, 1971 NUCLEOTIDE POOLS IN E. COLI SYNCHRONOUS CELLS 81 CTP, or a combination of high dttp and low dctp, interacts with a Psi factor for cell division. For example, dctp or CTP might inhibit transcription of division genes by interaction with a Psi factor analogous to ppgpp and Psi (ribosomal) during most of the cell cycle. At the end of a round of replication, when dctp or CTP is low, dissociation from the Psi (division) would occur and transcription of a group of distal division genes would ensue. ACKNOWLEDGMENTS We thank Jon Gallant and Barry Harada for teaching us the technique of measuring the nucleoside triphosphate pools and Clayton Bagwell who gave excellent technical assistance. We are very grateful to G. A. O'Donovan, J. Neuhard, and R. Werner for reading the manuscript and offering many helpful suggestions. This investigation was supported by grant 4419 from the National Research Council and by a Studentship from the Medical Research Council awarded to L.H. LITERATURE CITED 1. Adler, H. I., W. D. Fisher, and A. D. Hardigree Cell division in Escherichia coli. Trans. N.Y. Acad. Sci., p Brown, N. C., and P. Reichard Role of effector binding in allosteric control of ribonucleotide diphosphate reductase. J. Mol. Biol. 46: Clark, D. J Regulation of deoxyribonucleic acid replication and cell division in Escherichia coli B/r. J. Bacteriol. 96: Cleaver, J. E Thymidine metabolism and cell kinetics, p. 63. North-Holland Publishing Co., Amsterdam. 5. Edlin, G., and G. S. Stent Nucleoside triphosphate pools and the regulation of RNA synthesis in E. coli. Proc. Nat. Acad. Sci. U.S.A. 6: Eron, L., R. Arditti, G. Zubay, S. Connaway, and J. R. Beckwith An adenosine 3':5'-cyclic monophosphate-binding protein that acts on the transcription process. Proc. Nat. Acad. Sci. U.S.A Halmstetter, G. E., and D. Cummings An improved method for the selection of bacterial cells at division. Biochim. Biophys. Acta 8: Helmstetter, C. E., and S. Cooper DNA synthesis during the division cycle of rapidly growing Escherichia coli B/r. J. Mol. Biol. 31: Helmstetter, C. E., S. Cooper,. Pierucci, and E. Revelas. On the bacterial life sequence. Cold Spring Harbor Symp. Quant. Biol. 33: Irr, J., and J. Gallant The control of RNA synthesis in Escherichia coli. II. Stringent control of energy metabolism. J. Biol. Chem. 44: Johnson, R. A., and R. R. Schmidt Enzymic control of nucleic acid synthesis during synchronous growth of Chlorelia pyrenoidosa. 1. Deoxythymidine monophosphate kinase. Biochim. Biophys. Acta 19: Lark, K. G Variation in bacterial acid-soluble deoxyribotides during discontinuous deoxyribonucleic acid synthesis. Biochim. Biophys. Acta 51: Lark, K. G Regulation of chromosome replication and segregation in bacteria. Bacteriol. Rev. 3: Neuhard, J Studies on the acid-soluble pool in thymine-requiring mutants of Eschenchia coli during thymine starvation. III. On the regulation of the deoxyadenosine triphosphate and deoxycytidine triphosphate pools in Escherichia coli. Biochim. Biophys. Acta 19: O'Donovan, G. A., and J. Neuhard Pyrimidine metabolism in microorganisms. Bacteriol. Rev. 34: Painter, P. A., and A. G. Marr Mathematics of microbial populations. Annu. Rev. Microbiol. : Pastran, I., and R. Perlman Cyclic adenosine monophosphate in bacteria. Science 169: Randerath, K., and E. Randerath Ion-exchange chromatography of nucleotide mixtures on poly(ethyleneimine) cellulose thin layers. J. Chromatogr. 16: Randerath, K., and E. Randerath Thin-layer separation methods for nucleic acid derivatives, p In S. P. Colowick and N.. Kaplan (ed.), Methods in enzymology, vol. 1, part A. Academic Press Inc., New York.. Stubblefield, E., and S. Murphree Synchronized mammalian cell cultures. 11. Thymidine kinase activity in colcemid synchronized fibroblasts. Exp. Cell Res. 48: Travers, A. A., R. 1. Kamen, and R. F. Scheif Factor necessary for ribosomal RNA synthesis. Nature (London) 8: Turner, K., R. Abrams, and 1. Lieberman Levels of ribonucleotide reductase activity during the division cycle of the L cell. J. Biol. Chem. 43: Werner, R Mechanism of DNA replication. Nature (London) 3:57-57.